In an anthropological forum, there was once a view that because Latin, Greek, Sanskrit (also Celtic IIRC) are Mora-timed, they are divided into one subgroup. However, "syllable-timed" Iranian and "stress-timed" Balto-Slavo-Germanic are classified into another subgroup. That view also argues that languages may develop in the direction of Mora-Timed>Syllable-Timed>Stress-Timed, and Mora-Timed is closely related to long vowels.

However, we know that in the phylogenetic tree model, this is unlikely. First of all, Iranian languages (at least well-attested languages like Avestan, Persian, Pashto etc.) and Sanskrit must consist one certain subgroup. Secondly, the model that classifies Latin, Greek and Sanskrit into one subgroup, but Balto-Slavo-Germanic into another, except for one Schleicher's tree, most other trees are not like this. For example, in Thomas Olander's tree model, Italo-Celtic is the third to split after Anatolian and Tocharian, predating Germanic and Balto-Slavic.

Therefore, I have some doubts about the relationship between mora-timed languages, quantitative metre and long vowels.

  1. First of all, are mora-timed and long vowels directly related? Many modern Romance languages have lost the opposition of long and short vowels in Latin and PIE, so they have become syllable-timed. But Proto-Iranian, Proto-Balto-Slavic, and Proto-Germanic all retain the opposition of long and short vowels. The following linguistic forum also recognizes that Lithuanian is a Mora-timed language, just like classical Latin, Greek, Sanskrit and Japanese. Therefore, in Balto-Slavo-Germanic, are there more Mora-timed languages (whether historical or modern)?

  2. My second doubt is the relationship between Mora-timed and quantitative verse. According to Wikipedia, in addition to classical Latin, Greek, and Sanskrit, classical Persian, classical Arabic and Old Church Slavonic also use quantitative metre. However, Persian is a stress-timed language. On the contrary, although Japanese is a Mora-timed language, according to Britannica, Japanese often uses syllabic verse.

  • Leaving aside that similar features existing in different languages doesn't mean they have to be inherited from a common ancestor, drawing conclusions about language typology based on verse types is dangerous. Latin adopted Greek metres for cultural reasons, not linguistic ones, and they were always a poor fit.
    – Cairnarvon
    Jan 25, 2023 at 0:02
  • That Britannica article misdescribes Japanese moraic verse as syllabic verse: haiku are composed not of 17 syllables but of 17 morae. Jan 25, 2023 at 4:06

1 Answer 1


First, the division of languages into stress-timed, syllable-times and mora-timed is wishful thinking. Phonetic timing differences between languages are much richer than implied by this division.

Second, however, there is a difference between languages that distinguish heavy vs. light syllables (as opposed to ones where there is no phonological distinction). Amongst languages with a syllable-weight distinction, the primary distinction is between languages where long vowels and diphthongs define heavy syllables, versus languages where in addition coda consonants also define heavy syllables (and there can also be a distinctions in which consonants in the coda make a syllable heavy – sonorants, vs all consonants.

The concept "mora" is widely invoked as a device for making sense of this typology. First, it is presumed that the segmental concept "long" is to be represented with a distinctive mora (a vowel with two moras is long and a short vowel has only one; a long consonant has a mora and a short consonant has none). See Morén 1999 for discussion. That means that by definition, languages with long vowels have a moraic contrast. I suppose that the reasoning behind the claim about mora-timed vs. syllable-times languages relates to the historical loss of vowel length in parts of Indo-European. Retention of an old phonological feature is not evidence of a shared phonological innovation (that should be obvious) – perhaps the idea was that a change to syllable-timing (i.e. lack of moraic distinction in syllables) defines a historical innovation.

In those languages which make a distinction between heavy vs. light syllables, we find the following divisions. 1: There are long vowels, and only long vowels make syllables heavy. Coda consonants do not make a syllable heavy. 2: There are no long vowels, and only coda consonants can make a syllable heavy. 3: There are long vowels, and coda consonants make syllables heavy.

We generally simplify the analysis by saying that quantitative meter is based on distinguishing light vs heavy syllables, where a light syllable has one mora and a heavy syllable has two, but also noting that coda consonants can be moraic, perhaps by general rule ("all (sonorant) consonants in the coda are moraic").

  • Thank you! English still has both long and short vowels, but has it lost the distinction between light and heavy syllables, making classical meters such as Hexameter unsuitable for it? In addition, most ancient Germanic poems are also of stress-meter. Therefore, in poetry tradition, is there really a phylogenetic difference between Greek-Latin-Sanskrit and Balto-Slavo-Germanic? In the proto languages related to Iranian and Balto-Slavo-Germanic, the opposition of long and short vowels is well preserved. Lithuanian, Persian and OCS can also adapt to quantitative verse. Jan 24, 2023 at 17:47
  • 1
    There is some evidence from the phonology that English distinguishes heavy from light syllables, especially in the stress system. Yet another complication is that "heavy" may not be the same in grammar as it is in poetry. You're poking at the edges of an interesting historical question that I don't control but has been written on, namely comparative Indo-European poetic meter.
    – user6726
    Jan 24, 2023 at 18:16
  • 1
    Latin is a language whose recent history shows a move from being characterized by vowel length and a poetry based on moraic feet to a language with unpredictable stress, no vowel length to depend on, and a poetry based on stress and rhyme. The differences between Classical and Goliardic Latin are amazing.
    – jlawler
    Jan 24, 2023 at 18:34
  • @user6726 The point of view in the anthropological forum is actually marginal and impossible. In the view, it seems that Latin and Greek were derived from "Pre-Pre-Proto-Indo-Aryan", while Balto-Slavo-Germanic from "Pre-Pre-Proto-Iranian". Therefore, the hypothesis of the split of poetic meter is invoked. Jan 24, 2023 at 18:34
  • @user6726 Although such a branch tree model is indeed impossible, what I still need to confirm is that whether the affinity for quantitative verse had been largely reduced during the proto-language stages of Germanic and Balto-Slavic. This book tells us that Germanic and Balto-Slavic prefer tonic and syllabic verses rather than quantitative. academic.oup.com/book/9392/chapter-abstract/… Jan 24, 2023 at 18:52

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.